Actions are a core capability within Webex AI Studio, enabling the AI Assistant to not only suggest responses but also recommend or perform specific tasks. These tasks can be performed within the Webex Contact Center environment or through integration with external systems. As an administrator, you define these actions within Webex AI Studio, specifying their purpose, triggers, required user information, and fulfillment methods.

This article details the process of configuring actions for your AI Assistant skills, including defining action types, specifying user inputs, and configuring fulfillment modes to enhance agent efficiency and customer experience.

Understanding action types and fulfillment modes

Action type Description When to use Agent interaction
Moderated mode Requires agent review and explicit approval before execution. High-stakes actions, data sensitive, or when agent input or verification is required. Agent reviews collected information, can modify it, and then clicks Approve to proceed.
Unmoderated mode Executes autonomously without agent intervention. Routine, low-risk, or repetitive tasks where no agent input or verification is required. Agent receives real-time status updates and confirmation of the action’s completion.

Prerequisites

Before configuring actions for AI Assistant skills, ensure the following:

  • Webex AI Studio access: You must have administrator privileges to access the Webex AI Studio platform.
  • Existing AI Assistant skill: A skill must already be created in the Webex AI Studio. Actions are configured as components of an AI skill.

For more information, refer to the Create and manage AI Assistant skills article.

Configure actions for AI Assistant skills

To add and configure a new action for a skill, perform the following steps:

1

Sign in to Control Hub and navigate to Services > Contact Center.

2

Go to AI Assistant under Desktop Experience in the navigation pane.

3

Under the Suggested responses section, click the Manage AI Assistant skills link. The AI Assistant skills dashboard opens in a new browser tab within the Webex AI Studio.

4

Click the skill for which you want to configure actions.

5

Navigate to the Actions tab and click + New Action.

You can configure a maximum of 10 actions for a skill.

6

On the Add a new action page, specify the following details:

  • Action name: Enter a clear, concise, and descriptive name for the action. This name is displayed to agents. Maximum length is 75 alphanumeric characters.

    Examples: "Get Flight Details," "Create Case," "Schedule Callback."

  • Action description: Provide a brief explanation of the action’s purpose. This helps agents understand its function. Maximum length is 1024 characters.

    Example: "This action will use the Customer PNR and retrieve the flight details."

  • Action scope: Select how the action interacts with data and fulfillment:
    • Slot-filling: The action focuses on collecting specific information (slots) from the conversation, without external system fulfillment. Agents can then copy and paste this information into any third-party system to complete an action.
    • Slot-filling and fulfillment: The action collects information and then uses it to interact with external systems or workflows.
  • Slot filling > Input entities (user inputs): Define the information the AI skill needs from customers or other systems to perform the action.
    • If Slot-filling is selected, only the Input entities section is shown.
    • If Slot-filling and fulfillment is selected, both Input entities and Webex Connect Flow Builder Fulfillment sections are displayed.

      Configuration: Add input entities in a table format or by using the JSON editor.

      • Table Format: Click + New input entity and specify the entity name, type (example, String, Number, Custom List/Choices, Regex, Date, Time, Phone, Email, Digit), description, and examples. Mark an entity as Required if mandatory.

        Understanding Input Entity Types: When defining an input entity, you select a type that specifies the kind of data the AI Assistant skill expects to collect. Here's a breakdown of the available types:

        • String: For free-text input, such as a customer's name or a general query.
        • Number: For numerical input, like quantities or ages.
        • Custom List/Choices: For predefined selectable options, where you provide a list of acceptable values (example, "Small," "Medium," "Large" for a size).
        • Regex: For custom validation rules for specific input patterns, ensuring data adheres to a particular format (example, a specific product ID format).
        • Date: For collecting date information, often presented with a date picker in the agent interface.
        • Time: For collecting time information, often presented with a time picker in the agent interface.
        • Phone: For phone numbers, with built-in validation for common phone number formats.
        • Email: For email addresses, with built-in validation for standard email formats.

        Digit: For a fixed number of numerical characters (example, a 4-digit PIN).

        Configuring Fulfillment Mode:

        • Set Agent Review required for each input entity. This option specifies if the agent needs to review or modify the input before an action is performed. This setting directly determines whether the action operates in Moderated or Unmoderated mode: If set to Yes, the action is performed in the Moderated mode (requiring agent review and approval).
        • If set to No, the action is performed automatically in the Unmoderated mode.
      • JSON Editor: Click Use JSON instead to define the input parameter schema in JSON format. For more information on JSON schema structure, see A tour of JSON Schema.
  • Webex Connect Flow Builder Fulfillment: Defines the workflow for performing the action after it is triggered. For detailed instructions on configuring fulfillment flows in Webex Connect for AI Assistant skills, refer to the Configure Fulfillment Flows for AI Agent Actions article.
    • Select service: Choose the required service configured in the Webex Connect client workspace. Webex AI Studio can only access services and flows configured within the client-level workspace.
    • Select a flow: Choose the required flow configured in the Webex Connect client workspace.
7

Click Add to save the new action to the AI skill.

8

After configuring the action, click Publish on the AI skill configuration page to make the skill live and apply the changes.

By carefully configuring actions, you enable your AI Assistant to provide dynamic, actionable guidance—significantly enhancing agent capabilities and improving overall customer interaction efficiency.